Keshav Agarwal

Software Engineer

Hyderabad, Telangana, India9 yrs 9 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Expert in Data Analytics and C# programming.
Successfully reduced code base secrets from millions to zero.
Led projects optimizing data pipelines for Bing.

Stackforce AI infers this person is a SaaS-focused Software Engineer with strong data analytics capabilities.

Contact

Skills

Core Skills

Data AnalyticsC#Software Development

Other Skills

Azure CommerceData APICosmos DBEvent HubAzure SearchAzure SQLAzure Data BricksKnowledge GraphPipelinesOptimizationAzureMLFortifySecrets DetectionVSTSHystrix

About

Experienced Software Engineer with a demonstrated history of working in the computer software industry. Skilled in Data Analytics, and C# and having a strong desire for learning new technologies. Strong engineering professional with a Master of Science (MSc) focused in Mathematics and Computer Science from Indian Institute of Technology, Kharagpur.

Experience

9 yrs 9 mos

Total Experience

4 yrs 10 mos

Average Tenure

9 yrs

Current Experience

Microsoft

2 roles

Software Engineer 2

Sep 2021 – Present · 4 yrs 9 mos

AZURE COMMERCE
Reporting Data API
Explore MS
Cosmos DB
Event Hub
Azure Search
Azure SQL
Azure Data Bricks

Azure CommerceData APICosmos DBEvent HubAzure SearchAzure SQL+3

Software Engineer

Jun 2017 – Sep 2021 · 4 yrs 3 mos

BING----------------------------------------------------
★Part of Knowledge Graph Team(Satori) of Bing
⇨ Scaled up pipelines to power Knowledge Graph based experiences on Bing in 98+languages and 200+ regions
⇨ Worked on optimization of pipelines to reduce overall latency of the system
⇨ Incrementally improved experience by powering newer set of data for the users
Digital Security and Risk Engineering----------------------------------
★Project: Auto-triage findings produced by Fortify
⇨ Used AzureML to test multiple models on the manually triaged reliable dataset of 40000 findings
⇨ Impact: Around 25% of the findings are currently being triaged by the module developed.
Approximately 58000 findings have been triaged saving 10000 man hours of the developers
★Project: Develop functionality to detect and report secrets in code bases
⇨ Developed the functionality to download the complete code base of all the repositories present in VSTS/MicrosoftIT and run the tool automatically on each one and report the results per organization
⇨ Impact: The total secrets reduced from 3.7 million to ZERO within a span of 1 year.
⇨ Extended the tool to scan escrets in build and release pipeline. Devised a format which Cred Scan could scan and yield maximum results

Knowledge GraphPipelinesOptimizationAzureMLFortifySecrets Detection+2

Intuit

Summer Intern

May 2016 – Jul 2016 · 2 mos · Bengaluru Area, India

★Project: APPLICATION RESILIENCY
⇨ Objective was to make applications resilient so that any single service failure does not bring down the entire consumer system
⇨ Used Hystrix to isolate points of access to services and stop cascading failure in complex distributed systems where failure is inevitable
⇨ Logged different metrics such as command name, execution time to get more insight about Facility Condition Index and track progress
⇨ Worked efficiently and wrote around 1500 lines of code to complete the task well within the project deadline of January, 2017

HystrixMetrics LoggingSoftware Development

Rajendra prasad hall of residence

Captain

Jul 2015 – Apr 2016 · 9 mos · IIT Kharagpur

⇨ Groomed and led a team of 30+ players to secure Bronze medal in penultimate year
⇨ Initiated Video Analysis to create self-awareness of errors and managed INR 1.2 Lakhs for various facets of inter and intra-hall cricket
⇨ Pioneered intra-hall night cricket tournament to promote the spirit of sports among boarders; witnessed a participation of 160 players

Trendwise analytics

Summer Intern

May 2015 – Jul 2015 · 2 mos · Bengaluru Area, India

★Project: PREDICTIVE MODELING
⇨ Devised a model to predict the number of cabs ordered at a given location, within any given time interval and date in R statistical tool
⇨ Critiqued model forecasts the availability between any two given dates; helping to analyze the effect of days in trend of the cabs
⇨ Improved the model, using the Random Forest technique, to achieve an accuracy of ~ 93% and made it robust under all assumptions

Predictive ModelingRData Analytics